The goal of this study is to examine the effectiveness of an aggressive follow up nursing intervention for homeless persons receiving prophylactic treatment for infection with Mycobacterium Tuberculosis (latent M. Tb.). The specific aims are to i) test the effectiveness (completion rate, cost, acceptability) of a nursing intervention to maintain contact with homeless clients during their course of prophylactic treatment for latent M. Tb.; 2) determine which site, person, and method options for the intervention are chosen, how often each is used, and effectiveness of each during the course of prophylactic treatment; 3) describe nurses' rationale for specific choices within each component in the intervention in relation to nurse and client variables; 4) describe nurse variables that may affect selection of specific options within each component and their effectiveness; 5) describe organizational policies that may affect generalizability of the intervention; 6) describe client, volunteer locator, and shelter worker variables that may affect the success of the intervention; and 7) determine if selected client or nurse variables have an impact on completion of prophylaxis. Homeless populations are particularly vulnerable to infection with M. Tb. Although effective prophylactic treatment is available, it requires continuous contact and follow up for the 6 month period that medication must be administered. The only accurate way to know that the medications are taken is by direct observation at each time the medication is to be taken. However, with the homeless population contact for direct observation for the 6 month period is problematic. There are methods of maintaining contact with this highly mobile population. The intervention has 3 components with designated options per component that can be selected for use by the nurse. A quasi experimental design enrolling control subjects in phase 1 and experimental subjects in phase 2 will be used. Two groups of subjects, n = 100 per group. will be selected from those requiring 6 month prophylaxis for latent M. Tb. Those in the baseline group will receive regular care and those in the experimental group will receive regular care plus the intervention. At the end of the research period, the groups will be compared on rate of completion of prophylaxis; cost of nursing time spent per client and cost of travel for field work per client; the client, nurse, shelter worker, and volunteer locator variables that might affect the success of the intervention will be described; specific use of each of the components and the nurses' rationale for selection of the component with a specific client will be described; nurses, clients, shelter workers and buddy's perception of barriers and benefits to the use of the intervention will be described; and perceived organizational policies that may affect the outcome will be described.